AlphaCode: DeepMind takes on competitive programmers

In one sentence DeepMind unveils AlphaCode, a system that generates code for competitive programming problems and ranks in the top half of human participants on Codeforces.

Verified Official source

ShareLinkedIn X

DeepMind, Google's AI lab, tries to make a model solve Codeforces challenges, the online contests where programmers race to solve problems.

AlphaCode isn't as good as the top humans, but it beats half of the participants. This isn't "autocomplete": it has to read a problem in English, understand it and write an entire program that passes tests.

The takeaway is that AI is no longer just helping you type faster, it's starting to reason about what to write. From here on, the "copilot that writes the program" debate becomes very real.